Electronic Six-Image Pulse Diagnosis for TCM Syndrome Identification in Sleep Disorders (SIP-TCM)

This study was conducted at the Integrated Traditional Chinese and Western Medicine Hospital of Southern Medical University. It was designed as a prospective, single-center, observational diagnostic accuracy study to explore the relationship between objective pulse diagnostic parameters and traditional Chinese medicine syndrome differentiation in patients with sleep disorders.

Eligible participants were adult patients with sleep disorders who were able to complete pulse examination and questionnaire assessment and who provided written informed consent. Patients with acute disease, severe organic cardiovascular disease, anatomical abnormalities of the radial artery, pregnancy, or other conditions unsuitable for pulse signal acquisition were excluded.

Each participant completed a standardized Six-Image Symptom Questionnaire developed according to the traditional Chinese medicine zang-fu image framework. The questionnaire covered six functional image domains, including heart, liver, spleen, lung, kidney, and San Jiao or mingmen-related domains. Expert traditional Chinese medicine syndrome differentiation was independently performed by two senior traditional Chinese medicine physicians. If disagreement occurred, a third senior expert adjudicated the final predominant syndrome, which served as the reference standard.

Pulse signals were acquired using the KY-M-A1 three-dimensional electronic pulse diagnostic system. Bilateral pulse signals were collected from the cun, guan, and chi positions, yielding six pulse sites in total. Participants rested quietly before acquisition, and pulse signals were recorded under standardized conditions. The system extracted pulse waveform parameters, including pulse amplitude-related indices, to construct a Six-Image pulse vector.

The primary analysis evaluates pulse-syndrome concordance by comparing the Six-Image pulse vector with the symptom-based Six-Image vector using cosine similarity. Higher similarity values indicate stronger agreement between pulse-derived features and symptom-based syndrome domains. Expert-determined traditional Chinese medicine syndrome classification is used as the reference standard. Secondary analyses include the diagnostic consistency between instrument-based syndrome differentiation and expert assessment, as well as receiver operating characteristic curve analysis to evaluate the discriminatory performance of the electronic pulse diagnostic system.